Functional neuroimaging is increasingly enhancing our understanding of autism spectrum disorders (ASD) as disconnection syndromes, but because of its demands, most studies have focused on adults or older children. Conducting functional neuroimaging during natural sleep permits the study of young children and infants. Given the early onset of ASD, such studies may be crucial to the identification of biomarkers. However, the assumption that findings obtained during sleep can be generalized to wakefulness has not yet been systematically tested. This exploratory proposal represents a first step to fill this important gap by studying young children with ASD in two conditions - wakefulness and natural sleep - using resting-state functional magnetic resonance imaging (R-fMRI). By quantifying intrinsic functional connectivity (iFC) throughout the brain, R-fMRI provides a wealth of information about functional brain circuitry. While differences between asleep and awake R-fMRI in healthy adults have been described, to date, iFC measures in sleeping young children with ASD have not been systematically characterized, nor have they been contrasted with wakefulness. Accordingly, our overarching goal is to provide an initial systematic characterization of the stability (reliability) across states (awake and sleep) of whole-brain iFC n children with ASD. We propose to collect R-fMRI while awake and during natural sleep in at least 20 children with ASD between the ages of 66 to 90 months - starting at the youngest ages at which children can be successfully scanned while awake. We will survey whole brain iFC employing structural and functional parcellation units commonly examined in the literature. We will also compute other whole-brain voxel-wise measures of intrinsic brain functional architecture previously reported to be abnormal in ASD and which capture specific properties not otherwise characterized by traditional correlation analyses. These include Voxel Mirrored Homotopic Connectivity, Regional Homogeneity, Fractional Amplitude of Low Frequency Fluctuations, Degree Centrality and Independent Component Analyses. Our hypothesis generating aims are (1) to test for significant differences between wakefulness and sleep across children with ASD, and (2) to systematically characterize the stability (reliability) of between-individual differences in R-fMRI measures across states (asleep, awake) as indexed by intraclass correlation coefficients. We will also address additional questions regarding the effects of different parcellation systems on measures of stability, the stability of brain-behavior relationships with ASD measures, and derive initial estimates of within-session test-retest reliability. Finally, to maximize the impact of this effort, we will make fully anonymized data available to the scientific community every six months as the data are collected. We expect such prospective data sharing to further enhance the scientific value of the proposed efforts. This will accelerate the pace at which the collected data can be used as a foundation for future efforts to study increasingly younger children so as to delineate the underlying pathophysiology of ASD.